First, stop and read Camels and Rubber Duckies. It's an good overview of pricing strategies. This article assumes you've read that.

Joel derides market segmentation, and perhaps in the software world people do resent it - I know whenever I see "lite" and "pro" editions I know I want the pro edition for the lite price, but the examples he gives are of companies that do it extremely successfully.

The Gap's upscale and downscale versions do very well for themselves. Airlines also did market segmentation well for decades, it was only when they pushed the concept too far that people rebelled. $199 for coach, $1999 for first class indeed.

Perhaps he means "in a world where the only common ground between a geek and his mother is that they both know they can use the Internet for comparison shopping, segmentation is becoming harder and harder to do."

AAA & Senior Citizen discounts, are also successful examples of segemntation. If you're using your AAA discount, it's because you likely saw the "$19.99 per night" blinking neon sign outside, not because you showed up at the 5 star hotel looking to spend $1350 instead of $1400 a night.

Similar to the Gap - Banana Republic - Old Navy reationship is Toyata and Lexus. Or VW & Audi. Ford and Lincoln. Automobile manufacturers do this exceptionally with 3 or 4 different brands selling essentially the same car. Sometimes with different ammenities, sometimes with very little difference.

The Lexus team within Toyata does a lot of work to make sure everything you expect from an expensive car is there. They make it quieter, upgrade the electronics, fabric, trim and so on, so you are getting something for your money.

Another example of segmentation is sales on old models. There are those who seek them out, buying their summer clothes (gasp) in the summer because that's when they're selling off their stock to get ready for the fall lineup. Buying used books on Amazon rather than the full priced new editions, and so on.

I think Joel's just bitter because his "3 versions of Citydesk" pricing strategy didn't work. The problem there was that the mid-price version was clearly a dumbed down full version. Who wants a content management system with a 500 item limit?

Then he went after another version of segmentation with his "This version can only edit documents, not go in to design mode" edition. Again, that was clearly a full edition with a flag that didn't let you access all the features. Plus the price to get in was the same as the full edition. You paid $300 for the full edition, and the edit-only edition was $100, but you had to buy 3 minimum.

Perhaps what Joel should say is "Market segmentation doesn't work when the world finds out the low price version of your software is just the full price version with an article number inhibitor built in to it."

A great example of segmentation that we're seeing being played out right now in a really hilarious way is AOL vs. Netzero.

We all know the AOL ad with the mom who walks on the table and tells the board room everything she wants and the AOL guy says "okay" and the mom walks off happy.

And we all know the NetZero attack ad where the same thing happens but someone blurts out "You'd have to go to NetZero for that."

Well now there's a new ad where a woman goes to NetZero, and someone blurts out "I don't know, I use Netscape."

But wait a minute... I thought Netscape was AOL. So here they're differentiating on brand in order to price their competition out of the market. Without diluting the "AOL deserves a premium" brand, they can fire back at their competitor with a somewhat-recognizable brand called Netscape.

Another version of segmentation is CPU manufacturers. It costs just as much to make a 3,000,000,000 Mhz chip as it does to make a 0.0000001 Mhz chip, but they make both, and charge different amounts for both, and this strategy works for them.

The second half of Joel's Essay, I think he's been working on for a while. He's talked about the $1,000 - $75,000 price gap when selling to corporations before. That's more or less true. The limit on my corporate card without requiring approval from not just my manager, but my manager's manager was $500, so I had to buy the licenses for CityDesk one at a time on two seperate corporate cards.

He's also talked about nobody getting experience with your product because nobody used it in college (do people really use FogBugz in college?). There's some real truth to that, but in really large organizations the people making the decisions are B-School graduates who didn't use any software in college beyond MS Office and Napster, which is why Oracle and Solaris have a lock on mind share and a 250,000 person organization is told in no uncertain terms that that is what they have to use. Which is why I bought CityDesk, it was the only solution that didn't cost a million dollars. Even a free CM system was a million dollars in hardware, and we weren't allowed to use MySql anyway.

Selling to smaller companies without such formalized rules on spending is somewhat different.

The focus group stuff is half right. Most focus group operators know by now that the straight on approach doesn't work, so you bury the question you're really asking amidst a dozen others so you get a more off the cuff response. Or you ask kids which color walkman they want (black), but also give away free walkman's and see which one they take (yellow). So he's using the new breed of focus group's arguments against the old breed as an argument against focus groups in general.

Site licensing is a double edged sword. Winzip was a fun product to talk about a couple of years ago. Nobody owned a pay copy and everyone just clicked the shareware splash screen every time it launched, even though the button moved. Nobody knew how Winzip made money.

But Citigroup was standardized on it, and it came with every computer. With 250,000 employees that adds up. So did Citigroup pay $30 per license for a profit of $7,500,000 or did they get a company-wide license for, say $100,000.

And would Winzip have made the sale for $7.5 million?

I still like the references to Alice in Wonderland & The Simpsons. Nobody ever reference Alice in Wonderland in essays, unless they're about The Matrix and rabbitholes.

Discuss this article in the Forum.